Augmenting Language Models with Long-Term Memory

TL;DR

This paper introduces LongMem, a framework that enables large language models to memorize and utilize long-term context by decoupling memory retrieval from the core model, significantly enhancing long-context modeling and in-context learning.

Contribution

The paper proposes a novel decoupled architecture with a frozen backbone LLM and an adaptive side-network for long-term memory, enabling unlimited-length context handling and improved long-form memory in language models.

Findings

Outperforms existing long-context models on ChapterBreak benchmark.

Enlarges memory to 65k tokens for in-context learning.

Achieves significant improvements in memory-augmented in-context learning.

Abstract

Existing large language models (LLMs) can only afford fix-sized inputs due to the input length limit, preventing them from utilizing rich long-context information from past inputs. To address this, we propose a framework, Language Models Augmented with Long-Term Memory (LongMem), which enables LLMs to memorize long history. We design a novel decoupled network architecture with the original backbone LLM frozen as a memory encoder and an adaptive residual side-network as a memory retriever and reader. Such a decoupled memory design can easily cache and update long-term past contexts for memory retrieval without suffering from memory staleness. Enhanced with memory-augmented adaptation training, LongMem can thus memorize long past context and use long-term memory for language modeling. The proposed memory retrieval module can handle unlimited-length context in its memory bank to benefit various downstream tasks. Typically, LongMem can enlarge the long-form memory to 65k tokens and thus cache many-shot extra demonstration examples as long-form memory for in-context learning. Experiments show that our method outperforms strong long-context models on ChapterBreak, a challenging long-context modeling benchmark, and achieves remarkable improvements on memory-augmented in-context learning over LLMs. The results demonstrate that the proposed method is effective in helping language models to memorize and utilize long-form contents. Our code is open-sourced at https://aka.ms/LongMem.